Deep traps in In0.3Ga0.7As nFinFETs studied by Generation-Recombination noise

  • L. He
  • , H. Chen
  • , D. D. Guo
  • , L. N. Hu
  • , Y. Qin
  • , E. Simoen
  • , C. Claeys
  • , B. Kunert
  • , N. Waldron
  • , N. Collaert

Research output: Chapter in Book/Report/Conference proceedingsConference proceedingpeer-review

Abstract

The assessment of deep traps in In0.3Ga0.7As nFinFETs by Generation-Recombination (GR) noise spectroscopy is described in this paper. The gate voltage dependence of the corner frequency is studied for several devices with different gate lengths. Both gate-voltage-independent and gate-voltage-dependent corner frequencies are found. It is shown that the noise type turns from GR to Random Telegraph Signal (RTS) when the device size scales. For GR noise, the activation energy of the traps is extracted, which can tentatively be attributed to O-related traps and dislocations, with an effective density of these deep levels varying from 1011/cm2 to 1012/cm2. For RTS, the capture and emission time constant are extracted. It could be ascribed to a single trap in the gate oxide rather than in the film, and it is concluded that the trap level is lower than the Fermi level when no bias is applied.

Original languageEnglish
Title of host publication2017 International Conference on Noise and Fluctuations, ICNF 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781509027606
DOIs
Publication statusPublished - 19 Jul 2017
Externally publishedYes
Event2017 International Conference on Noise and Fluctuations, ICNF 2017 - Vilnius, Lithuania
Duration: 20 Jun 201723 Jun 2017

Publication series

Name2017 International Conference on Noise and Fluctuations, ICNF 2017

Conference

Conference2017 International Conference on Noise and Fluctuations, ICNF 2017
Country/TerritoryLithuania
CityVilnius
Period20/06/1723/06/17

Keywords

  • deep level traps
  • GR noise
  • InGaAs FinFETs
  • RTS

Fingerprint

Dive into the research topics of 'Deep traps in In0.3Ga0.7As nFinFETs studied by Generation-Recombination noise'. Together they form a unique fingerprint.

Cite this